Development of a Novel Approach for Controlling and Predicting Residual Sugars in Wines

Round 1

Reviewer 1 Report

This manuscript propose a novel model for predicting the residual sugar of the wine at a specific density point during fermentation, by measuring the initial brix of the must before fermentation. The model can be used as a practical and rapid tool for estimating the exact point in the fermentation at which to terminate the fermentation in order to achieve the requested residual sugars in the wine, thus gaining precision and saving time and costs. However, I recommend that the author address the following issues in order to improve the quality of the manuscript.

(1) The table is not standardized, please use three line tables.

(2) What specific improvements should the authors consider regarding the methodology? What further controls should be considered?

(3) There are many formatting issues in the literature.

(4) At present, the methods for measuring sugar, including glucose, are already relatively fast and simple. Why does the author still need to build a new model, and the advantages have not been reflected?

(5) Is the initial pH requirement for fermentation quite strict? Why choose a fermentation starting pH of 3.0? If the fermentation starting pH is not 3.0, do we need to redo the standard curve?

Minor editing of English language required.

Author Response

We thank the reviewer for his review.

Attached please find (in red color font) a detailed answer for all the reviewer’ comments.

1) The table is not standardized, please use three line tables.

All tables were standardized to three line tables.

(2) What specific improvements should the authors consider regarding the methodology? What further controls should be considered?

The specific improvements includes elaboration of the fermentation conditions to include the effects of different yeast strains, various temperatures and aeration conditions.  See lines 28-30.

(3) There are many formatting issues in the literature.

The formatting issues has been fixed.

(4) At present, the methods for measuring sugar, including glucose, are already relatively fast and simple. Why does the author still need to build a new model and the advantages have not been reflected?

Although there are already relatively fast and simple methods, most of them require special instruments (i.e. various sensors or enzymatic kits etc.) and with the model, there is no need for a special instrument rather than a hydrometer or refractometer, which exist in all wineries including the little ones.

(5) Is the initial pH requirement for fermentation quite strict? Why choose a fermentation starting pH of 3.0? If the fermentation starting pH is not 3.0, do we need to redo the standard curve? 

The fermentation starting point for the synthetics musts were 3.3. Most winemakers aim for a pH of 3.3 -3.6 for red wines. We chose the 3.3 to maintain a more stable wine and protect against bacteria.

Reviewer 2 Report

The authors present an attempt to predict the residual sugar concentration in fermenting wine by applying a correction formula to simple hydrometer density measurements. While the idea is great and has the potential to save time and money for winemakers, the study design lacks rigor and needs extensive validation work before it should be considered for publication.

General comments

The overall language is not scientific enough for the journal and needs to be revised. The use of "we"-statements should be the exception, however, they can be found throughout the manuscript.

Specific comments

Line 18   The residual sugar does not determine the wine style. A wine style is something like sparkling wine or Port wine. The residual sugar determines the category within a style like dry and sweet for still wines or brut for sparkling wines.

Line 24   Any type of legally binding information would require method approval by the OIV or TTB and I doubt that an indirectly calculated number would meet those requirements. Please be careful with these over-arching statements and promises of what a method can do.

Line 26   That is one of the major flaws in this study. The authors ignore the influence of the yeast strain on alcohol conversion rates. This is probably the most important factor and the calculator is based on a single strain's performance.

Line 36   This is factually wrong. Reducing sugars are not fermentable sugars. All fermentable sugars are reducing sugars but not vice versa.

Line 39   This is also incorrect. Residual sugar and reducing sugar is not used interchangeably.

Line 49   Mellow is not a legal term for wine.

Line 53   As pointed out above, style is not determined by residual sugar.

Line 56   Please define Brix here in the context of wine. It is important to understand that winemakers assume all brix to be sugar even though that is not accurate.

Line 61   This is not correct. The measurement is not inaccurate, sugar is just not reflected the same way. This whole correlation or lack thereof needs to be explained a lot better.

Line 66   The section here is not well organized. There is too much back and forth between methods which results in redundancies and confusion. Please reorganize.

Line 91   Please add a reference for that statement.

Line 96   I do not think that environmentally friendliness is a limiting factor for HPLC. Compared to the titration method which is using heavy metals, HPLC is very sustainable.

Line 102  Manufacturer information needs to be referenced correctly.

Line 105  How is hydrometry a destructive technique?

Line 130  Sugar concentration in fortified wines is very different and I do not think that specialty wine styles should be included in this work as long as the calculator is not properly validated for all of them.

Line 143  Please add supplier information for all your chemicals.

Line 155  This is a very central problem of this study. All your data is just based on one yeast strain. There are over 200 commercial yeast strains for wine on the market and they all differ significantly in their metabolic behavior and alcohol conversion rates. It is a bold over-simplification to assume that this model can be robust enough to work for all strains.

Line 156  What volume was the yeast added to? 10 g is not appropriate as addition information without context.

Line 163  Again, what volume were the 10 g added to? Also, why did you use a different strain for real wine? It makes the two experiments less comparable.

Lien 168  Where did the potassium metabisulphite come from and is 60 ppm the level of free or total SO₂?

Line 177  Referencing the OIV method number is enough. Please delete the "also appears in the OIV..." part.

Line 181  This is by no means a validation. Figure 1 is completely unnecessary as the WineScan is a recognized method and does not require verification. Please be careful when using these terms. You only verified data, you did not validate anything.

Line 198  Please reference the Microsoft Excel version properly.

Line 207  The first paragraph is entirely unnecessary. Please delete.

Line 214  Felling should be Fehling. Also the remark about weekends is strange in a scientific paper. The graphs show the missing data points.

Line 218  There are statistics missing in this part and in Figure 2. Please run statistics on your data to find significant differences.

Line 221  "Some residual sugar" is too vague. Please be more specific.

Line 234  The label is missing on the x-axis here. Also, differentiate between Figure 2a and 2b to make the discussion easier to follow.

Line 257  Why are you not showing all regressions?

Line 265  The axis label needs to be on the outside. Also, the R² in your paper all have a different number of significant digits. Please use the same number of digits throughout, which is typically 3.

Line 282  The R² should be added to the table.

Line 307  This is definitely not true. the winemaker needs a lot more data about more yeast strains and more winemaking conditions to be able to trust this calculation.

Line 333  All equations need to be numbered.

Line 349  Stopping a fermentation is a totally different issue. It requires a lot more information that a calculated sugar concentration or density. The manuscript is very theoretical without further validation data and the authors should not make these broad claims about winemaking applications without this data in hand.

Line 363  Please delete the phrase in brackets. Everything should be described in the methods and material section, so this is redundant.

Line 370  Please reformat the table, so it does not take up so much space and follows the template guidelines.

Line 422  Your worked with synthetic must, not synthetic wine.

Line 462  Please adjust the decimals for R².

Line 472  Why don't you focus on the accurate calculation from a sugar concentration from a density reading. The entire application for winemaking is only the next step and is taking a lot of focus away from the idea of this paper. I would rather like to see more yeast strains, more actual validation data, and better statistics than this calculator that winemakers can supposedly use.

Line 552  Please remove Figure 6. It does not add any scientific value to the manuscript.

Line 560  The first few sentences are simply repeating information from an earlier section. This is all redundant and not a real discussion.

Line 577  The application does by no means enable the user to achieve an accurate measurement. It would, given proper validation, be an estimate based on a calculation, not a measurement.

Line 583  The discussion about influencing factors is still ignoring the yeast strain, which is arguably the most influential factor.

Line 600  Looking at all the things you mention here would be the first part of the validation.

Line 603  This is not an appropriate conclusion for a scientific paper and needs to be adapted to the actual outcome of the study.

There are only minor edits required as far as grammar is concerned. The scientific writing styles requires a little more work.

Author Response

We thank the reviewer for his detailed and comprehensive review which improved the manuscript greatly.

Attached please find (in red color font) a detailed answer for all the reviewer’ comments.

General comments

The overall language is not scientific enough for the journal and needs to be revised. The use of "we"-statements should be the exception, however, they can be found throughout the manuscript. The language was revised and the “we” statements were omitted from the manuscript.

Specific comments

Line 18   The residual sugar does not determine the wine style. A wine style is something like sparkling wine or Port wine. The residual sugar determines the category within a style like dry and sweet for still wines or brut for sparkling wines. The sentence was revised at lines 18-19.

Line 24   Any type of legally binding information would require method approval by the OIV or TTB and I doubt that an indirectly calculated number would meet those requirements. Please be careful with these over-arching statements and promises of what a method can do. The sentence was revised to avoid over estimation of the method capabilities; see lines 24-30.

Line 26   That is one of the major flaws in this study. The authors ignore the influence of the yeast strain on alcohol conversion rates. This is probably the most important factor and the calculator is based on a single strain's performance. We rephrased the impact of our model, as abovementioned (lines 24-30).

Line 36   This is factually wrong. Reducing sugars are not fermentable sugars. All fermentable sugars are reducing sugars but not vice versa. Indeed, we fixed the mistake.

Line 39   This is also incorrect. Residual sugar and reducing sugar is not used interchangeably. We omitted the sentence.

Line 49   Mellow is not a legal term for wine. The whole paragraph was omitted from the manuscript.

Line 53   As pointed out above, style is not determined by residual sugar. The sentence was revised at line 45.

Line 56   Please define Brix here in the context of wine. It is important to understand that winemakers assume all brix to be sugar even though that is not accurate. Brix was defined in the text at lines 50-51.

Line 61   This is not correct. The measurement is not inaccurate, sugar is just not reflected the same way. This whole correlation or lack thereof needs to be explained a lot better. The sentence was revised at line 55.

Line 66   The section here is not well organized. There is too much back and forth between methods which results in redundancies and confusion. Please reorganize. The section was reorganized at lines 59-63.

Line 91   Please add a reference for that statement. We added references at line 81.

Line 96   I do not think that environmentally friendliness is a limiting factor for HPLC. Compared to the titration method which is using heavy metals, HPLC is very sustainable. We quoted Texeira et. al. (2022); nevertheless, we omitted this statement.

Line 102  Manufacturer information needs to be referenced correctly. We corrected the manufacturer information at line 92.

Line 105  How is hydrometry a destructive technique? It is destructive to fruits but not liquids. We omitted the paragraph from the text.

Line 130  Sugar concentration in fortified wines is very different and I do not think that specialty wine styles should be included in this work as long as the calculator is not properly validated for all of them. We omitted the wine styles from the manuscript.

Line 143  Please add supplier information for all your chemicals. Supplier information for all chemicals was added at line: 123.

Line 155  This is a very central problem of this study. All your data is just based on one yeast strain. There are over 200 commercial yeast strains for wine on the market and they all differ significantly in their metabolic behavior and alcohol conversion rates. It is a bold over-simplification to assume that this model can be robust enough to work for all strains. We reduced our statements regarding the model capabilities through the whole manuscript. 

Line 156  What volume was the yeast added to? 10 g is not appropriate as addition information without context.   We revised to concentration values (lines 136 and 144).

Line 163  Again, what volume were the 10 g added to? Also, why did you use a different strain for real wine? It makes the two experiments less comparable.   We revised to concentration. Regarding the yeasts, we used fx-33 with the synthetic must as the medium resembled white wine which this strain is used for fermenting.

Line 168  Where did the potassium metabisulphite come from and is 60 ppm the level of free or total SO₂?  Total.

Line 177  Referencing the OIV method number is enough. Please delete the "also appears in the OIV..." part. Deleted.

Line 181  This is by no means a validation. Figure 1 is completely unnecessary as the WineScan is a recognized method and does not require verification. Please be careful when using these terms. You only verified data, you did not validate anything. We deleted the figure.

Line 198  Please reference the Microsoft Excel version properly. Done at line 166.

Line 207  The first paragraph is entirely unnecessary. Please delete. Deleted.

Line 214  Felling should be Fehling. Also the remark about weekends is strange in a scientific paper. The graphs show the missing data points. We revised accordingly (line 159).

Line 218  There are statistics missing in this part and in Figure 2. Please run statistics on your data to find significant differences. We run statistics on all points; the points that looks like they are “missing” statistics have very small values of STDEV and thus appear missing. We revised the graph in order to emphasize better these points.  

Line 221  "Some residual sugar" is too vague. Please be more specific. We revised to the actual content at line 191.

Line 234  The label is missing on the x-axis here. Also, differentiate between Figure 2a and 2b to make the discussion easier to follow. We added the label to the x-axis and differentiated the figures.

Line 257  Why are you not showing all regressions? Because putting all of them in the same graph creates a blur of points that interfere with seeing the trend; all of the regressions are with high R² and presented in table 1.

Line 265  The axis label needs to be on the outside. Also, the R² in your paper all have a different number of significant digits. Please use the same number of digits throughout, which is typically 3. The axis is now outside and we revised the R² values through the whole text to 3 digits.

Line 282  R²  should be added to the table. We added R² to the table.

Line 307  This is definitely not true. the winemaker needs a lot more data about more yeast strains and more winemaking conditions to be able to trust this calculation. We revised the statement through the whole manuscript.

Line 333  All equations need to be numbered. Equations were numbered.

Line 349  Stopping a fermentation is a totally different issue. It requires a lot more information that a calculated sugar concentration or density. The manuscript is very theoretical without further validation data and the authors should not make these broad claims about winemaking applications without this data in hand. As mentioned above, we revised the statement.

Line 363  Please delete the phrase in brackets. Everything should be described in the methods and material section, so this is redundant. We deleted the phrase.

Line 370  Please reformat the table, so it does not take up so much space and follows the template guidelines. We reformatted the tables according to the template.

Line 422  You worked with synthetic must, not synthetic wine. We revised to synthetic must.

Line 462  Please adjust the decimals for R². The decimals were adjusted.

Line 472  Why don't you focus on the accurate calculation from a sugar concentration from a density reading. The entire application for winemaking is only the next step and is taking a lot of focus away from the idea of this paper. I would rather like to see more yeast strains, more actual validation data, and better statistics than this calculator that winemakers can supposedly use. One of the goals of was to develop a practical tool for the winemaker. Indeed, further development with more yeast strains, and variables should be done.

Line 552  Please remove Figure 6. It does not add any scientific value to the manuscript. Figure 6 was removed.

Line 560  The first few sentences are simply repeating information from an earlier section. This is all redundant and not a real discussion. These first few sentences serves as an introduction to the followed paragraph stating various studies for a comparison.

Line 577  The application does by no means enable the user to achieve an accurate measurement. It would, given proper validation, be an estimate based on a calculation, not a measurement. We revised the statement (lines 489-493).

Line 583  The discussion about influencing factors is still ignoring the yeast strain, which is arguably the most influential factor. We added the yeast strain to the influencing factors (lines 538-543).

Line 600  Looking at all the things you mention here would be the first part of the validation.

Line 603  This is not an appropriate conclusion for a scientific paper and needs to be adapted to the actual outcome of the study. We revised the conclusion section (lines: 552-555).

Reviewer 3 Report

I find the topic of this manuscript interesting to a group of readers involved with wine making on a commercial scale, where incorrect length of the fermentation process can result in considerable financial losses. The Authors proposed a novel application of relatively cheap ane easily accessible equipment; their approach is well-described and clearly explained.

Author Response

We thank the reviwer for his comments.

The reviewer had no comments that needs to revise. 

Round 2

Reviewer 1 Report

All the questions I raised have been addressed.

Author Response

We thank the reviewer for his comments.

All the questions he raised have been addressed.

Reviewer 2 Report

The revision fixed a few of the major flaws, however, there are still issues with the content of this manuscript that need to be addressed before it should be considered for publication.

Specific comments

Line 18  The legislation application is already mentioned above. This sentence is redundant.

Line 24  This is not a real sentence. Please rephrase.

Line 28  Future work should not be in the abstract. Please list a few more actual results instead.

Line 40  The introduction is still too fragmented. Please avoid all the separate paragraphs and keep topics together.

Line 43  How can fermentation time be adjusted? 

Line 48  I am still not convinced that stopping a fermentation is the best application for this analytical technique. It is in fact quite challenging to arrest an active fermentation even when one knows the target density.

Line 52  You are referencing the same source in two sentences in a row. Please revise.

Line 92  Manufacturer information is still not referenced correctly in multiple passages of the manuscript. Please be consistent: Manufacturer, City, State, Country.

Line 108 This sentence is redundant as you just mentioned the model development above.

Line 126 Please add a unit to the brix listings.

Line 134 Use repetitions instead of repeats.

Line 136 The correct unit is g, not gr.

Line 140 Should be ripening, not ripping.

Line 142 Use repetitions instead of repeats.

Line 143 Please avoid the use of brackets throughout the manuscript. In this spot, just use sugar level instead of treatment.

Line 149 You answered my question whether this was free or total SO2, but you did not add that information to the manuscript.

Line 153 Where are those indicated time points?

Line 159 Again, you did not do a validation and since you are not referencing the data anywhere, this part can be removed from the manuscript.

Line 166 Which Excel version did you use?

Line 220 The legend needs to be explained or it needs to show the units.

Line 288 This is not true for any given initial brix level, only for levels within the range that was tested.

Line 312 What does relatively high mean? Relative to what?

Line 319 Equations should be labeled as Equation 1, Equation 2, and so on. This way equations can be referenced correctly instead of saying "as follows".

Line 343 Everything is described in methods and materials, so this statement is redundant.

Line 350 This table still takes up too much space. If you manage to get parameters and their units into one line, it will be easier to see and cut the table size in half.

Line 406 Please check throughout the manuscript. Tables and Figures need to be capitalized.

Line 435 The labels on the y axis in Figure 4a need to be consistent in terms of decimals.

Line 448 Why 30 g/L residual sugar?

Line 520 Saccharomyces needs to be capitalized.

Line 539 Saccharomyces needs to be capitalized.

Line 551 It is unclear how this approach saves time or effort. I believe the cost savings but I am not sure about the other claims.

The overall text still requires some editing but especially the parts that were added during the revision need extensive language editing.

Author Response

We would like to thank the reviewer for his thoughtful comments and efforts towards improving our manuscript.

The revision fixed a few of the major flaws, however, there are still issues with the content of this manuscript that need to be addressed before it should be considered for publication.

Attach please find, in red font, the answers to the issues raised.

Specific comments

Line 18  The legislation application is already mentioned above. This sentence is redundant. The sentences were omitted.

Line 24  This is not a real sentence. Please rephrase. The sentence was rephrased. Lines 22-29.

Line 28  Future work should not be in the abstract. Please list a few more actual results instead. The future work was omitted, and actual results were added.

Line 40  The introduction is still too fragmented. Please avoid all the separate paragraphs and keep topics together. The paragraphs were joined together.

Line 43  How can fermentation time be adjusted? This sentence was replaced by a full description of the methods for fermentation cessation. Lines 40-42.

Line 48  I am still not convinced that stopping a fermentation is the best application for this analytical technique. It is in fact quite challenging to arrest an active fermentation even when one knows the target density. We think that premature cessation of fermentation is a common practice in the wine industry, leading to the production of a wide range of products. The availability of a method to calculate accurately the residual sugar at a given density point can therefore be very useful when fully developed.

Line 52  You are referencing the same source in two sentences in a row. Please revise. We revised.

Line 92  Manufacturer information is still not referenced correctly in multiple passages of the manuscript. Please be consistent: Manufacturer, City, State, Country. Manufacturer information was revised through the text, accordingly.

Line 108 This sentence is redundant as you just mentioned the model development above. The sentence was omitted.

Line 126 Please add a unit to the brix listings. The word was corrected.

Line 134 Use repetitions instead of repeats. We revised accordingly.

Line 136 The correct unit is g, not gr. The unit was corrected.

Line 140 Should be ripening, not ripping. The word was corrected.

Line 142 Use repetitions instead of repeats. The word was corrected.

Line 143 Please avoid the use of brackets throughout the manuscript. In this spot, just use sugar level instead of treatment. We revised accordingly.

Line 149 You answered my question whether this was free or total SO2, but you did not add that information to the manuscript. We added the information. Line 143.

Line 153 Where are those indicated time points? We added at line 146 the words- as indicated at Figure 1.

Line 159 Again, you did not do a validation and since you are not referencing the data anywhere, this part can be removed from the manuscript. The part was removed.

Line 166 Which Excel version did you use? We added the version of excel. Lines 154-155.

Line 220 The legend needs to be explained or it needs to show the units.  Units were added to the legend.

Line 288 This is not true for any given initial brix level, only for levels within the range that was tested. Line 288 was revised.

Line 312 What does relatively high mean? Relative to what? We erased the term relatively high and added the value. line 279.

Line 319 Equations should be labeled as Equation 1, Equation 2, and so on. This way equations can be referenced correctly instead of saying "as follows". We labeled accordingly.

Line 343 Everything is described in methods and materials, so this statement is redundant. We revised the statement and left only the essential part.

Line 350 This table still takes up too much space. If you manage to get parameters and their units into one line, it will be easier to see and cut the table size in half. The table was revised and the size cut.

Line 406 Please check throughout the manuscript. Tables and Figures need to be capitalized. Tables and Figures are now capitalized throughout the whole manuscript.

Line 435 The labels on the y axis in Figure 4a need to be consistent in terms of decimals. The labels were revised and consistent.

Line 448 Why 30 g/L residual sugar? Just an example, in the range of semi-dry wines.

Line 520 Saccharomyces needs to be capitalized. Saccharomyces was capitalized.

Line 539 Saccharomyces needs to be capitalized. Saccharomyces was capitalized.

Line 551 It is unclear how this approach saves time or effort. I believe the cost savings but I am not sure about the other claims. We believe that for small wineries which do not have expensive instruments, measuring density and brix compared to measurements of residual sugars is faster and requires less effort.

Comments on the Quality of English Language

The overall text still requires some editing but especially the parts that were added during the revision need extensive language editing. The English was revised.